Count every qualifying Top 8
Only complete Top 8s from tournaments matching your filters contribute. Each appearance earns points; five are required for a rank.
Research workspace / Overview
Stored on your computerPublished results · updated dailyTHE FORMAT, IN FOCUS
Know the field.
Explore the decks, results, and patterns behind Premodern.
AT A GLANCE / CURRENT SETTINGS
The current top five
Loading rankings…
Collecting results
ARCHETYPE LANDSCAPE
Share of the Top 8
FROM THE FIELD TO THE FINAL
Progression at a glance
THE SHORTLIST
Archetype finish breakdown
One result, one finish: 1st, 2nd, 3rd-4th, or 5th-8th.
Archetype statistics
Finish columns are exclusive: a winner counts only in 1st. Click a column to sort or an archetype to inspect results.
WHY THIS ORDER?
Achievement points you can trace.
Reward deeper finishes
1st = 3 points, 2nd = 2, 3rd-4th = 1.5, and 5th-8th = 1. These already include the Top 8 reward.
Reward larger tournaments
Multiplier = 1 + 0.5 × log₂(players / 16). Add finish points × multiplier for every result. Unknown attendance uses ×1 and is marked below. There is no 100-point cap.
Switched off in the current view: every result counts ×1.
These explanations are calculated from saved standings. Card art represents each archetype; Rogue uses Eron the Relentless as a mascot. Unassigned images await a card choice. Review proposed archetype groupings ↗
Every deck, explained
Open a deck for its position, finish breakdown, calculation and source results.
Card art © Wizards of the Coast, LLC. Images via Scryfall; artist credits appear within each deck explanation. This app is not affiliated with Wizards of the Coast.
OUTSIDE THE ARCHETYPE RANKING
Rogue & unclassified decks
Every published result under Rogue, Other, Unknown, Unclassified, or another combined source label in your current selection. MTGTop8 names every deck, so its Rogue decks are the ones whose name no other MTGTop8 deck uses; each still shows its MTGTop8 name.
These labels may contain different deck builds, so they do not receive an archetype rank. Their results remain in the statistics and eligible Top 8 totals. Open a decklist to inspect the individual build. Decks whose saved list matches an established archetype of either source (80%+ of the main deck; 70% between Survival decks) are counted with that archetype instead and are not listed here, as are "Survival - Other" decks and unnamed Survival decks with 25+ creatures, which count as Survival.
Event library
Source records, deck coverage, and collection status.
Excluded from analysis ? Parallax Tide
Collection history
TOP 8 SHARE BY MONTH
Who is rising, who is fading
Share of all Top 8 slots each month, in events with a complete Top 8. Choose archetypes below; hover or use the arrow keys for exact values.
Show as a table
FORMAT HEALTH
How diverse are the Top 8s?
Effective number of archetypes per month: 1 / Σ share². Higher means more decks share the Top 8 slots. Hollow points: months with fewer than five events.
Show as a table
WHERE THE EVENTS ARE
Regions
FROM THE DECKLISTS
Most-played cards
AGGREGATE DECKLIST
Average list
Each card copy is ranked by how many lists play at least that many; the most common copies fill 60 main-deck and 15 sideboard slots.
READ THE NUMBERS CORRECTLY
Coverage comes before conclusions.
TC Decks and MTGTop8 usually publish the top finishers, not the whole field. A page with 8 decks from a 100-player tournament is that event's Top 8: it shows how those eight players finished, but not what the other 92 registered. Field Lab therefore keeps the two apart: Top 8 statistics use published Top 8s, while metagame share uses only events whose complete field is published.
Exclusive finish counts
1st | 2nd | 3rd-4th | 5th-8thEach recorded tournament result enters exactly one finish column. A winner adds one to 1st and zero to the other finish columns. Top 8 appearances is a separate total for all placements 1-8, not an additional finish or points award. Rank ranges crossing these bands are reported as unresolved.
Share of recorded decks
recorded decks of this archetype / all recorded decksDescribes the available sample. It can overrepresent successful decks. It is not a full-field metagame estimate.
Metagame share
archetype field entries / all field entriesUses only events with every unique exact placement from 1 through the reported player count and a complete Top 8. Events with fewer than 8 players are excluded from this cohort.
Share of Top 8
archetype Top 8s / all Top 8sUses events with all eight unique exact Top 8 placements recorded. MTGTop8 also accepts a complete set of elimination bands: one 1st, one 2nd, two 3rd–4th and four 5th–8th. Unknown attendance never establishes a complete field. Total recorded finishes may also include incomplete events; these counts are kept separately.
Top 8 → 4 → 2 → 1
Top 4 / Top 8; Top 2 / Top 4; Top 2 / Top 8; wins / Top 8; wins / Top 2Uses the same complete-Top-8 cohort for every stage. These are placement progression rates; final standings alone do not prove that playoff matches were played.
Representation index
Top 8 share / metagame shareBoth shares use the same complete-field cohort. A value of 1.5× means the archetype's Top 8 share is 1.5 times its field share. Sample size and event selection still matter.
How reliable is a rate?
95% Wilson interval of hits / appearancesConversion rates come with the range the true rate likely lies in, given the sample size. ▲ / ▼ mark rates that are reliably above / below what a random Top 8 slot reaches: 50% to Top 4, 25% to the final, 12.5% to the title. A deck that won 3 of 7 Top 8s shows 43%, but its range is roughly 16–75%.
Trends, diversity and regions
effective archetypes = 1 / Σ (archetype share of Top 8 slots)²Monthly shares use the Top 8 slots of complete Top 8 events. The effective number of archetypes is 20 when the Top 8s are as varied as twenty equally common decks. Regions come only from locations stated in event names; MTGTop8 names most locations, TC Decks fewer. Months with fewer than five events are drawn hollow.
Card analysis
average list: card copies ranked by how many lists play themUses every saved full decklist in the current selection. Core cards appear in at least 90% of an archetype's lists, common cards in 50–90%, flex cards in fewer. The average list takes, for each card, the copies most lists agree on, until 60 main-deck and 15 sideboard slots are filled.
Inferred Top 8 match win rate
inferred wins / (inferred wins + inferred losses)Off by default. Confirm a conventional 8-player single-elimination bracket in an event's details to enable it. First = 3–0, second = 2–1, third/fourth = 1–1, fifth–eighth = 0–1. Do not confirm events with byes, intentional splits, concessions or a third-place match. This is an estimate, not observed match data.
Automatic archetype matching
shared main-deck card copies / larger main deck ≥ 80%Decks labelled Rogue, Other or Unknown, and labels used only once, are compared card by card (lands included, sideboards excluded) with every saved list of established archetypes. A deck joins an archetype when at least two of its saved lists match 80% or more and no other archetype with two such lists comes within 5 points (a single mislabelled source list does not block a match). When two archetypes do come that close, the source is usually naming one deck two ways (TC Decks "4c Control" and "Nic Fit"), so the deck joins the one with at least twice as many matching lists; if neither has, it keeps its label. Survival of the Fittest toolbox decks run many one-ofs, so a deck playing Survival joins a Survival archetype from 70%. A deck that matches nothing in its own source is compared the same way with the other source's lists (TC Decks with MTGTop8 and back). It then takes the other source's name, or this source's name when at least 75% of the other archetype's lists match it (MTGTop8 "RG Aggro" is TC Decks "Zoo"). A Survival deck that still matches nothing counts as "Survival" when the source labelled it "Survival - Other", or when it is unnamed (Rogue) or has a one-off name containing Survival and plays Survival of the Fittest with at least 25 creatures (card types from Scryfall). A deck the source labelled as Survival keeps that in its name: a "Survival - Other" list that matches Elves is counted as "Survival - Elves", not with plain Elves. MTGTop8 never labels a deck Rogue; it gives each one a name, even a one-off brew. An MTGTop8 deck whose name no other MTGTop8 deck uses and whose list matches no archetype is therefore counted as Rogue, like TC Decks' unnamed decks (or as Survival when its name says Survival). The original label stays visible as the source label; other labels are never changed.
Source fidelity & exclusions
Labels that differ only in case, punctuation or word order are grouped under the source's most used spelling ("Survival Welder" and "Welder-survival" count as "Welder Survival"); other archetype labels stay as listed by the source. Missing or ranged ranks are retained in raw exports and excluded from strict coverage checks. 5–0/undefeated league reports are excluded by default and never treated as tournament wins. Unknown event sizes are excluded when a size minimum is set.
Parallax Tide: post-ban exclusion
The official ban took effect 18 January 2026 (20 January on Magic Online). TC Decks collection searches both maindeck and sideboard card indexes across the post-ban range, then verifies every match in the actual decklist. Any confirmed match excludes the entire event, including events after 25 January. Transition-week events (18?25 January inclusive) stay out of analysis until this audit completes. Absence from the card index means no matching deck was found in the source; decks absent from the source cannot be checked. MTGTop8 collection checks every published deck’s main and sideboard list and saves an audit per event. Source pages and exclusion evidence are retained.
Collection & reproducibility
Collection uses a real local browser, saves each event atomically, and waits between requests. Saved results do not expire. Completed date ranges and Tide audits are recorded, so reruns collect missing or failed events and check only the last seven days of listings for new arrivals. Older late-posted events can be found with the optional historical listing check. A SQLite backup is created before every run. Re-running skips previously collected events unless “Force refresh saved results” is enabled. Collection always saves all event sizes and sources within the collection dates, including league reports. The attendance slider, source selector, and analysis dates only filter your saved data; changing them never requires re-scraping already collected events.
CSV exports include raw counts; the JSON research bundle also includes filters, collection history, event coverage, and generation time. Raw source snapshots are saved in data/pages. All data stays in this folder on your computer.This website shows a read-only copy of the collected results. Collection runs automatically once a day; visitors cannot start collection or change the data, and the filters only change what you see.